Experiments with AI Code Review
Engineering Blog ● Covered by 2 sources
Wealthfront built an AI code review system called Iris that uses Claude Opus as the lead agent with adversarial sub-agents (prosecution and defense models) to evaluate potential bugs and issues in pull requests. The system achieved an average review cost of $4 and 10-minute turnaround, with engineer ratings skewing toward useful and great feedback while significantly reducing false positives compared to earlier attempts. The AI review complements rather than replaces human code review, running after self-review and before peer review to maintain reviewer independence.
Why it matters
Wealthfront went on a multi-year experiment to improve AI-augmented code review processes, transitioning from initial skepticism to a structured approach that uses models like Opus 4 and Gemini for effective bug detection and code analysis. Their iterative refinements have led to improved review outcomes and reduced false positives.